November 18, 2004
Contact: chris riggall
Secretary Cox: New Data Shows Voting Accuracy Soared in
ATLANTA … Secretary of State Cathy Cox today released newly compiled data from Georgia’s November 2nd General Election showing that the accuracy of the vote count increased dramatically compared to the presidential election four years ago. Secretary Cox said the improvement was attributable to the deployment of a uniform electronic voting system in 2002.
Secretary Cox said the improvement in accuracy resulted in more than 103,000 additional Georgia ballots being properly counted in the presidential race this year.
In the presidential election in 2000, Georgia’s undervote rate (sometimes called the residual vote rate) which is the percentage of ballots that registered no choice in the presidential race, was 3.5% of ballots cast. Academic researchers with the CalTech/MIT Voting Technology Project found that Georgia’s undervote rate in 2000 was the second-worst in the nation.
Final data from the 2004 General Election reveals that the presidential undervote rate in this year’s election plummeted – dropping nine-fold, to only 0.39%. Out of more than 3.3 million ballots cast this year, only 12,843 ballots registered no choice for president.
“This new data makes clear that the adoption of a modern system of electronic voting dramatically improved the accuracy of the vote count in Georgia,” said Secretary of State Cox, who serves as the state’s chief elections official. “If you apply the high undervote rate we experienced four years ago to the huge turnout we had this year, we would have had more than 103,000 additional lost votes in the presidential election in 2004. In other words, modernizing our voting system resulted in 103,000 more Georgians having their presidential choice counted this year. To improve voting accuracy by such a huge margin may be unprecedented in American electoral history. Four years ago, no state, whatever their voting method, had an undervote rate as low as 0.39%, the number we recorded on November 2nd. Georgia has gone from a laggard to a leader in voting accuracy, and our citizens now have the assurance that the votes they cast will be accurately recorded and counted,” Ms. Cox added.
Researchers indicate that, even in a presidential contest using the most accurate voting equipment, some small number of voters will intentionally abstain from selecting a candidate. While achieving a statewide undervote percentage of zero is not a realistic goal, a modern touch screen voting interface that prohibits overvotes (improperly selecting more than one candidate) and offers users a review screen so they can see and, if necessary, correct their choices before a ballot is cast, provides voters a much better opportunity to reduce inadvertent error. Votes tabulated electronically are also not subject to tabulation errors that can occur with punch card and other paper-based voting systems.
The November 2nd General Election was the fifth major statewide election to utilize the uniform statewide electronic voting platform, which was first deployed for the November 2002 General Election. In addition, several hundred local, municipal and special elections have been conducted using the touch screen terminals.
Opinion polling has consistently shown that Georgians are more confident in the voting process today than they were before the deployment of electronic voting. In exit polling conducted Nov. 2nd for the broadcast networks, CNN and other major news organizations, Georgians were shown to be above the national average, and significantly above the average for southern states (+5%), in expressing confidence that their vote would be accurately counted. Floridians, for example, were 17 percentage points more likely than Georgians to say they were not confident that their votes would be counted.
Other points of interest from the new data that compares undervotes in 2000 to 2004:
· Four years ago the county with the highest undervote percentage in the state was Randolph, with a rate of 15%, followed by Bacon County with a rate of 10.5%. In 2000 some 56 Georgia counties recorded an undervote percentage of 5% or higher.
· This year, the county with the highest percentage was Taliaferro, with a rate of 2.26%, followed by Hancock at 1.77%. Those higher-than-average rates were primarily attributable to paper absentee ballots, the data shows. In 2004, 62 Georgia counties recorded undervote rates of 0.5% or below.
· In 2004 the county with the lowest undervote percentage was Forsyth, at 0.12%. Only 66 Forsyth ballots out of 56,985 cast in the county showed no choice in the presidential race.
· The undervote rate among voters who cast ballots via touch screen on Election Day was only 0.35% -- a bit lower than the overall percentage of 0.39%. Among absentee and early voters, (which includes a substantial number of paper optical scan ballots) the rate was 0.52%. For provisional ballots, which are always cast on paper optical scan ballots, the presidential undervote rate was 1.16%.
A study of Georgia’s deployment of electronic voting and its impact on the 2002 election, released last month by Dr. Charles Stewart of MIT, one of the nation’s foremost experts on voting system accuracy, noted,
“Following the 2000 presidential election, the state of Georgia instituted the most comprehensive overhaul of voting technology in the country…Although the Nation’s eyes were on Florida, a case could be made that it was Georgia that deserved the scrutiny. Georgia’s “residual vote rate,” a measure of “lost votes” that has come to be used widely to measure voting technology reliability, was 3.5% -- the second worst in the country, behind only Illinois…This paper shows that the implementation of the Diebold system produced a significant reduction in the residual vote rate throughout the state of Georgia. Just as important, the implementation of the new machines removed gaping disparities in voting machine reliability that could have raised serious questions about the fairness of Georgia’s electoral system. DREs are more reliable, their performance varies less across the state, and the least advantaged areas of the state have experienced the greatest gain in reliability.”
Note to Editors: Attached is an Excel spreadsheet that provides county-by-county details on ballots cast, presidential votes recorded, and the difference between those two in both real and percentage terms (the undervote) for the 2000 and 2004 General Elections in Georgia.
The first tab, 2004 Presidential Analysis, provides a thorough review of the 2004 vote. The columns from left to right show:
1) (yellow columns) The overall undervote,
2) (green columns) The undervote for voters who cast their ballots in person on Election Day using touch screen electronic terminals,
3) (brown columns) The undervote among those who cast absentee ballots, including advance voters. Note that these ballots may be cast via paper optical scan ballot or by touch screen.
4) (white columns) The undervote among those who cast an optical scan paper provisional ballot on Election Day.
The second tab, Comparison, 00 vs 04, provides a county-by-county comparison of the overall undervote rates in 2000 and 2004, and the percentage improvement for each county.
The third tab, Scatterplot US President, depicts in graphical form the distribution of county undervote percentages in the 2000 and 2004 elections.
3,317,336 ballots were cast in this year’s election. Georgia’s undervote rate four years ago was 3.5%. If that rate had prevailed this year, 116,106 presidential undervotes would have been recorded. The number of undervotes actually recorded in 2004 was only 12,852, an improvement of 103, 254.